College Football Predictions

The Harvard College Sports Analysis Collective (HSAC) released predictions on Sunday for what they thought the AP Poll would look like when released on Monday morning. And their prediction model did fairly well:

Our predictions did well, accurately predicting the exact ranking of 11 teams. That is not counting minor mis-orderings, like the Wisconsin-Kansas State-Oklahoma triad. We struggled a bit at the bottom of the poll, as we suspected.

Finally, the correlation between the predicted points for each team and the actual points was an amazing 0.985.

These are pretty impressive results given that the AP Poll is based on somewhat subjective human voting behavior. Since this is just one random week taken from the "population," I would be even more interested to see how well HSAC's predictions hold up over an entire season. I've commented on their post with this question, and am currently waiting for a response.

My intuition is that their predictions would be less accurate the first half of the season as the best teams start to have not yet proven themselves in voters' minds (voters being sportswriters and coaches). Toward the end of the season, I would guess that these predictions would have greater precision as teams have largely sorted themselves by that point and voters have more data (wins, loses, points per game, opponent points per game, etc.) to base their rankings on. This is simply a hypothesis, but worth looking into.

HT John M. for sharing this post